What are symbols symbols of? Situated action, mythological bootstrapping and the emergence of the Self

Abstract

Abstract:  This paper addresses the question of how symbols should be understood in analytical psychology and psychoanalysis. The point of view examined focuses on the recent turn to more cognitive and developmental models in both disciplines and briefly reviews and critiques the evolutionary and cognitive arguments. The paper then presents an argument based on dynamic systems theory in which no pre-existing template or structure for either mind or behaviour is assumed. Within the dynamic systems model the Self is viewed as an emergent phenomenon deriving from the dynamic patterns existing in a complex system that includes the physiological characteristics of the infant, the intentional attributions of the caregiver and the cultural or symbolic resources that constitute the environment. The symbol can then be seen as a discrete, and in important ways an autonomous, element in the dynamic system. Conclusions are drawn for further research into the nature of the symbol with implications for both theory and practice in analytical psychology and psychoanalysis.

Introduction¹

In a recent publication, Jean Knox, our colleague in this conference, warned of the dangers psychoanalysis and analytical psychology face should they neglect to take into consideration ongoing advances in cognitive science and neuropsychology (Knox 2002). I take it that Jean's sense of danger served as at least one stimulus behind the title of this conference. How are we to reconcile the traditional issues of the depth psychologies with developments taking place in the other sciences of mind? At the last international congress of analytical psychologists, in a debate concerning the relationship between Jungian thought and evolutionary theory, it was my place to warn that we do not fall prey to the problem Freud faced late in life when it had become clear to his closest followers that his adherence to Lamarckism, the inheritance of acquired characteristics, was untenable, and posed a serious scientific threat to psychoanalysis. Freud's rejoinder, ‘so much the worse for the biologists’, does not embody a position we want to emulate.

But there are serious problems, even dangers, in the other direction as well. Freud's remark reflected, as Patricia Kitcher (Kitcher 1992) has magnificently demonstrated in her book, Freud's Dream, the fact that the theoretical difficulties Freud encountered at an increasing rate after the turn of the century originated in his embrace of an array of 19th Century scientific propositions that were themselves on the verge of radical transformation. Thus, prior to 1900 and the rediscovery of Mendel's genetics, even leading Darwinians, such as George Romanes, whose book, Mental Evolution in Man, significantly influenced Freud (Sulloway 1979), were committed Larmarckians, as in fact Darwin himself had been. Freud tied significant elements of his system to Larmarckian Darwinism, only to see it collapse with the rise of neo-Darwinism, which linked natural selection to Mendel's gene model.

The problem that we face in responding to the theme of this conference, and to Jean's altogether appropriate warning, therefore, is that when one surveys the present state of the sciences of mind one does not find consensus. Indeed, I doubt there are many other scientific fields where dissention is as intense, and widespread, as it is in the fields to which we would look for the new paradigms we are seeking. I have spent several years now examining the most triumphalist of the new paradigms, evolutionary psychology, and have concluded that it is fraught with a host of methodological and conceptual difficulties that are probably intractable. Evolutionary psychology, despite its claim that it has finally connected psychology to the other natural sciences by way of its evolutionary pretensions, is in fact far more dependent on the so-called cognitive revolution of the 1950s and 60s, building on the work of Turing, Chomsky, Fodor, and others, than it is on neo-Darwinism. And even there, one of the originators of the cognitive model, Jerry Fodor, has recently taken the evolutionary psychologists to task on their three most important arguments, massive modularity, computation, and the evolution of mind (Fodor 2000).

To make matters worse, the cognitive paradigm, as it has existed for half a century, is now under attack from several quarters. The manifest failure of strong computational artificial intelligence models to generate naturalistic behaviour has resulted in a sea change in computer modelling, most notably in the rise of autonomous agent research, more commonly known as situated robotics. Situated robotics unquestionably constitutes a paradigm shift, in Kuhn's sense, and its leading theoretical feature, first proposed by Rodney Brooks in the 1980s (Brooks 1999), has been to banish the cognitive processor from the modelling process and move directly from perception to patterns of action. Valentino Braitenberg had previously proposed this model in more abstract terms, and the simplest instantiations of the robotic paradigm can still be referred to as Braitenberg Machines (Braitenberg 1984).

Even before Brooks jettisoned the cognitive module from his robots, traditional AI had come under attack from another advance in computer theory, neural network or connectionist modelling. This paradigm shift derived from advances in massively parallel computation in which it became possible to model a simple nervous system. With this method, researchers such as Jeffrey Elman and Luc Steels quickly began to undermine one of the basic premises of the cognitive revolution, Chomsky's assertion that grammar had to be innate because it was too complex to be learned. Neural network modelling proceeds with little or no programming, but Elman, Steels and others were nevertheless able to teach machines basic grammatical patterns in relatively short time (Elman et al. 1996; Steels & Vogt 1997). The connectionist model was also incorporated into psychiatry, where it has provided important insights into psychopathology in the hand of researchers like Manfred Spitzer at the University of Ulm in Germany (Spitzer 1992; 1999). More traditional cognitivists, such as Steven Pinker, struck back, and the debate rages on (Pinker 1997). Similarly, connectionists, such as Andy Clark, work mightily to pull the radicals from robotics and dynamic systems theory back to something resembling the traditional analytic methods the neural network paradigm still embraces (Clark 1998).

And then there is dynamic systems theory. DST has much in common with situated robotics, although it comes closer to the interests of depth psychology in so far as several of its strongest advocates are researchers in the field of infant development. DST is in most regards a truly developmental theory, focusing on the notion of emergence. It is also the case, no doubt in large measure due to this developmental orientation, that DST, or the general notion of emergence, has caught the attention of psychoanalysts such as Robert Stolorow (Stolorow 1997), Stephen Mitchell (Mitchell 2000), and Beatrice Beebe and Lachmann (Beebe & Lachmann 2002), and Jungian analysts including our colleagues at this conference, Joe Cambray, Jean Knox, and David Tresan. But there are issues that we must confront before we rush to embrace DST. Most important among the issues are the radical implications of strongly emergent models. In a recent paper in the Journal of Analytical Psychology, Patricia Skar, a Jungian analyst, and Peter Saunders, a mathematician and theoretical biologist, proposed what is in effect an inversion of the traditional manner in which Jungians think about the relationship between the archetypes and the complexes, arguing that it was in the formation of complexes that the archetypes appeared as emergent properties (Saunders & Skar 2001). This argument is based in a dynamic systems mode of thinking, but I do not believe most Jungians have recognized its radical nature. By the same token, in reading Stolorow's embrace of DST, I have come away with the sense that he does not realize that, for example, the work of Ester Thelen and Linda Smith (Thelen & Smith 1998), which he relies on heavily, necessarily undermines major assumptions in psychoanalysis, even to the point of calling into question the notion of the self, an issue to which I will return below.

The upshot of all this is that we cannot simply argue that we must connect our analytic theory to the advances in the other sciences of mind. There is not ‘an other’ science of mind, there are many – indeed, the list of controversies could be multiplied many fold beyond where I have left it, and I have not even touched on the controversies in neuroscience and neuropsychology. Much has been learned in all of these fields, but chaos and confusion abound as well. This means that if we are to look for new paradigms for understanding our own discipline we will need to have some principled means of sorting out the alternatives. In order to do this I will turn to the work of Horst Hendriks-Jansen, whose book, Catching Ourselves in the Act, is to my mind one of the most important studies of the entire field of theories of mind recently published (Hendriks-Jansen 1996). Hendriks-Jansen has the added virtue of having directly addressed topics of interest to clinicians, specifically the relationship of infant research to the autonomous agent theory of situated robotics, and a probing critique of cognitivist models of autism (Hendriks-Jansen 1997).

To anticipate, Hendriks-Jansen is an advocate for the situated robotics/dynamic systems theory models of mind, and critic of the computational/modularity models. The question then is whether there is a principled basis for embracing this point of view rather than the other potential candidates. Hendriks-Jansen's argument is that there is, and that it derives from the history of modelling. Problems, from Hendriks-Jansen's point of view, originate from misleading assumptions about the nature of modelling and the nature of mind that derive from a perverse fascination with the computational model, beginning with Turing, as early as 1936. To illustrate this argument, think for a moment about one of the most profound modelling insights in the history of medicine, William Harvey's insight that the heart was a pump. What Harvey did was draw an analogy between a relatively new technology, the valved pump, and the heart. You had a machine that moved fluids with a series of imbedded mechanisms, and you could draw instructive analogies to the machine that moved blood in the body. The functional structure of both machines was sufficiently analogous to allow for the definition of what philosophers of science call ‘natural kinds’, such as the valve in the heart. One could also draw distinctions – the number of chambers – that yielded new insights. As Hendriks-Jansen argues, Harvey's insights followed a long tradition of modelling based on analogies between objects in the world and human capabilities or organs.

To abbreviate a complex argument, Hendriks-Jansen maintains that the computer, whether in its traditional AI or in its connectionist manifestations, does not occupy the same analogical space as Harvey's pump. As Hendriks-Jansen puts it, the computer is analogically inert, because it can be used to simulate anything. Proper programming can make the same machine give you either a simulation of the pumping of the heart or the movements of the stars. But either simulation requires, as Turing was the first to recognize, a complete functional description, or set of parameters, in order for the simulation to take place. What one can learn from such a simulation is what happens when you vary a parameter. In a computer based simulation of the heart, for example, one might learn what would happen if a valve lost some of its elasticity, but one would first have to know that the heart contained elastic valves that normally operated within certain parameters.

But why would one embrace robots or dynamic systems for modelling purposes? The simple answer is that both of these approaches rely first of all on mechanical operations in natural environments. Their foundations rest in the same territory as Harvey's pump. In the case of robotics, the foundations are found in actually functioning machines performing tasks such as wall following, cleaning up spaces, or co-operating to move a large object. Dynamic systems theory, despite its complex mathematical formulation, rests on the mechanical behaviour of weighted springs and pendulums. For Thelen and Smith, for example, the development of a motor skill such as grasping can be analysed using the same sorts of differential equations used to analyse the movement of a weighted spring. We can therefore model the infant's arm and grasping behaviour as analogous to the movement of a weighted spring. But what about mind? Here things do become more complex, and it is with the problem of mind that I want to move into the real issues that I believe confront us in looking at new paradigms.

Situated action

I have already noted that dynamic systems theory carries with it some critical issues for modelling. The first issue we confront with both robotics and DST is that neither of them starts out with a model of mind. Mind has to be arrived at by way of models of action – situated action to be more specific. This is at the heart, for example, of the inversion perpetrated by Skar and Saunders in their paper on archetypes. Ever since Jung first proposed his theory of archetypes in 1919 Jungians have worked on the assumption that there had to exist some place – in the brain, in the genes, in some Platonic realm of ideas – a template for the archetypes. Jung himself was clear that this was his assumption, and he drew a distinction between what he called the archetype in itself – the template – and the archetypal image. To Jung, and to his successors, this was the only way in which you could account for the observation of common patterns of ideation and imagery in myths and other patterns defined as archetypal. Susan Oyama, in her critique of innatist models in biology, has argued that this notion that a plan has to exist before a thing can exist is a common prejudice in Western thought, beginning with Plato (Oyama 2000). Jung is clearly party to that prejudice. Similarly, this is the case with Freud and his successors. One of the most important problems Freud had to solve early on was how to constitute the unconscious in the neonate in such a way that repression, which requires both a push from above and a pull from below, could take place. His solution was the equivalent of original sin in the form of what he called primal repression. It was here that his Lamarckism became so important, because this ‘plan’ for the unconscious had to be acquired within historical time – hence the myth of the primal horde.

In dynamic systems theory, however, no plan for grasping is necessary in order to achieve grasping behaviour. All you need, Thelen and Smith argue, is a set of springs (muscles), attached to an armature (skeleton), that move a set of weights (baby fat). Crucially, however, you have to locate that machine in the environment for which it was selected by evolution, the environment of adults and artefacts that we call the human world. Evolution is vital to both the robotic and the DST models of mind, but not the evolutionary story told by evolutionary psychology, or even those of Freud and Jung. The robotics/DST point of view is that evolution builds simple machines that can develop rapidly and reliably into adult systems if they are in an appropriate environment. That environment is not the Pleistocene savannah, it is the world of adult conspecifics already well on their way through the developmental space of the species. As Thelen and Smith demonstrate, it is the mother, observing the arm movements of her child, who places objects within the range of the infant, and thereby helps to entrain muscular coordination that leads to adult grasping. Similarly, as Kaye and Wells have demonstrated, the human infant has a unique sucking pattern that becomes the object of interpretation by the mother leading to a coordinated pattern of interaction which, they speculate, is a precursor of conversational turn taking (Kaye & Wells 1980). In these and other instances, the mind that first engages the pattern of action is the mother's adult mind, attributing intentionality to the infant.

Attribution, however, introduces another aspect of dynamic systems that I believe is crucial to our understanding of developmental processes, the dynamic linking of time frames. As Kelso, Ding, and Schoner make clear, the ways in which varying time frames intersect often drives the formation of patterns in dynamic systems (Kelso, Ding & Schoner 1993). This can take several forms. At its simplest, the developmental pattern of a given system is rarely linear. Thus the phase transitions in heated fluids used by Skar and Saunders to illustrate their argument are characterized by relatively long periods of pattern stability even while heat increases, followed by abrupt transitions to new patterns, which again display stability. In complex interactive systems, one encounters interlocking time frames that dynamically influence one another. Relevant to our discussion, for example, is the real time process of infant-caregiver interaction as it interacts with developmental time. One issue that I want to suggest derives from this way of thinking is that we can posit and investigate a series of nested time frames, and their dynamic interaction. One can, then, speak about real time, nested in developmental time and on through cultural time to historical and finally to symbolic and mythic time. As I hope to show, viewing the symbol as a part of a dynamic system, particularly in relation to the temporal dynamics of a system, will illuminate a number of important issues for the depth psychologies.

Mythological bootstrapping

So how do we get a mind, a psyche, a self out of a dynamic systems model? Let me first put on the table a qualification of the modelling process that brings us back, to a degree, to modules and computation. One cannot maintain that springs and weights, or simple robotic perception-action cycles alone yield a human mind. Let me be very clear about this: Thelen and Smith argue that dynamic systems modelling indeed shapes brain structure, following the work of Gerald Edelman (1987), such that as the developmental process takes place the brain is configured to carry out these processes on an increasingly organized and systematized basis. Similarly, in the field of situated robotics, Rolf Pfeifer at the University of Zürich has demonstrated that a robot that starts out using simple sensors to navigate a room can pass off its parameters to a neural network, which will, over time, take over the navigation process, displaying such features as anticipation of known objects and other developmentally sophisticated behaviour (Pfeiffer & Scheier 1999). Nevertheless, the foundation of this kind of mind rests on the situated action of the machine, and its relationship to the environment. In neither case is there a pre-existing template in the system.

It is also important to understand that this position is not just some variant on Skinnerian behaviourism. It is not based on learning theory. Rather the process by which mind is achieved in these models is bootstrapping. The essential factor here is that the infant is adapted by evolution to display patterns of action that the human adult caregiver has come to understand as meaningful. This is, we might say, the myth of intentionality. Intentionality is at the heart of psychoanalytical thinking, and owes its central place, I believe, to Freud's studies with Brentano, a subject that requires far more attention than it has received. Brentano's argument was that all acts of consciousness were intentional – in the philosophical sense – in that all acts of consciousness have an object. They are directed toward something. Freud extended this argument to the notion that there was unconscious ideation. If there are patterns of action to which the individual attributed no conscious meaning, that is no intentionality, then, by way of the hypothesis of the unconscious, intentionality could be recovered. The intentional object of the action was unconscious. In fact, Freud's and, indeed, Klein's theoretical move was not that far removed from the ordinary adult response to infantile patterns of action. The attribution of intentionality, both in the pragmatic sense of wanting to do something and in the philosophical sense of having an object of consciousness is the habit of mind that is engaged by the adult caregiver in his or her interactions with the infant. By attributing intentionality – in both senses – to a pattern of action, one attributes mind to the infant. Intentionality implies that the actions of the infant are about something. However, the somethings about which the attributions are made are in fact not about anything in the DST model. They are first of all evolved patterns of action that have been selected because they fit into the dynamic interactive system known as human development. This notion, which I will now elaborate, is at the heart of the question posed in the title of this paper; what are symbols symbols of?

I can illustrate what I mean by the somethings about which attributions are made by reference to an important simulation, published in the proceedings of the second international conference on artificial life, carried out by Edwin Hutchins and Brian Hazelhurst of the University of California at San Diego (Hutchins & Hazelhurst 1990) following on research by Ackley and Littman (Ackley & Littman 1992). Their simulation set about to model the processes by which a culture could acquire the capability to predict an event that no individual could learn to predict in a single lifetime. The case in point was the need to predict tidal variation in order to successfully move a foraging tribe from an inland area to the sea in time to harvest molluscs. This required that a population acquire the means to correlate the phases of the moon with the pattern of the tides – a distinctly temporal problem I might note. In the course of the simulation it was demonstrated that over a prolonged evolutionary time frame, it was possible to derive a population that was somewhat better at innately predicting the tides. In other words, simple evolution will get you part of the way. Much greater success was achieved, however, when all of the agents in a given generation produced an artefact that could be used to predict the tides. The next generation was then set to select the best artefacts and attempt to improve on them. This is a form of what is called Baldwinian evolution, after the eminent developmental psychologist, James Mark Baldwin who, along with the comparative psychologist, Conway Lloyd Morgan, first proposed it in 1896 (Baldwin 1896/1996; Hogenson 2001). In Baldwinian evolution, a complex process of ‘co-evolution’ is initiated involving both the organism and the organism's environment. In the case of the tide prediction simulation, the evolution of the artefacts was at least as important as the evolution of mental abilities in the agents. As Hutchins and Hazelhurst make clear, the notion of artefacts that evolve can be applied to virtually any cultural form, including myths, and we might say fantasies.

The notion that cultural artefacts, including myths, enjoy some meaningful form of evolutionary development is increasingly accepted in philosophical and anthropological research. A notable example is the German philosopher Hans Blumenberg, who has argued that myths display a kind of evolutionary quality, what Blumenberg calls ‘Darwinism in the realm of language’ (Blumenberg 1985). The anthropologist and neuro-scientist Terrance Deacon has also proposed that language in general, but we might add specifically those stories that last, i.e., myths, has evolved to work within the brain and developmental setting of the human infant (Deacon 1997). Needless to say, the increasingly popular notion of the meme, despite serious problems of definition and theoretical coherence, follows the same pattern (Blackmore 1999). In all, it seems to me that rather than think in terms of myths deriving from some primordial, transcendent, or genetic blueprint, we should view myths and symbols, as Hutchins and Hazelhurst explicitly suggest, as discrete artefacts that interact dynamically with other elements in a complex system of agents, environment and other artefacts. From this point of view, the argument from DST would then be that myths and symbols are part of the system that bootstraps the infant, and subsequently the developing individual, into the world of intentional objects, meaningful action, and relationships.

The emergence of the Self

At the risk of gross over simplification, I want to assume a critical position vis à vis the history of theorizing about the self based on the following disjunctions: non-linear versus linear processes, including time frames; state transitions through phase spaces versus stage transitions; attractor formation versus trauma; and emergence versus development. Time does not allow an extended discussion of these disjunctions, but I believe they can be made to fall out fairly naturally, at least for present purposes. Beginning at least with Freud, human psychological development has been largely conceived of in linear terms. A series of stages were postulated, through which normative development was believed to progress, baring any interference with the progression. Any such interference would lead to some degree of failure in the progression of the individual's development. Such interference could be called by various names, but we have tended to dramatize such events by calling them traumas. Trauma results in a failure to complete some aspect, however small, of the developmental process, and one observes these events in such phenomena as the neuroses, regressions, splitting, and other forms of psychopathology.

To proceed in an equally abbreviated fashion with a dynamic systems approach to the question of the emergence of the self, the first claim made is that a theory of stages is nothing more than a series of arbitrary cuts through a process that in fact has no intrinsic structure. Competencies, which tend to make up the substance of a stage, are essentially attributions on the part of individuals (adults) who have successfully formed a basin of attraction around a particular task. What is missing from the stage model is the realization that the infant, or indeed the adult working through Erickson's stages of life development, is in most instances not really working on achieving the alleged goal. What is in fact going on is movement through a phase space, which can be thought of by analogy to the movement of a pendulum through the range of its swing. The range of the swing, in turn, defines a basin of attraction for the movement of the pendulum. Any perturbation of the pendulum alters the phase space, and in turn defines a new attractor. What is seen as a stage in classical theory is seen as a relatively stable attractor in dynamic theory. But by moving from a stage theory to a notion of attractor states within phase space we will have to rethink such reliable old analytic workhorses as the notion of regression. As Michael Tucker and Kathryn Hirsh-Pasek remark, ‘apparent developmental regression’ results from an ‘increased systemic variability at phase transitions in the dissolution of one stable attractor as the system moves toward another’. ‘From an objective standpoint’, they continue,

the system appears to become less complex, more disorganized. The system is also more sensitive to disruption or trajectory changes at these points. However, following this brief, variable period, the system will reorganize, and the ‘missing’ behaviors may spontaneously reemerge. Usually they will be more stable, reliable, and more complex than before the reorganization.

(Tucker & Hirsh-Pasek 1993, p. 366)

Another example using pendulums will illustrate this point. If one pendulum is attached to a wall, and set in motion, it will define a stable attractor within phase space, making a transition at each end of its swing. If a second pendulum is attached to the same wall, and set in motion, the two will first fall into what appears to be a more disorganized pattern, but eventually they will fall into coordination with one another. What is happening is that both pendulums create slight vibrations in the wall that interact with one another at the point of the phase transitions, leading to a chaotic regime and then to synchronization of their movement.

Now think for a moment, in an admittedly stylized fashion, about the process of analysis. When do we, as analysts, make judgements about interpretation? I would propose, that if we could map the interaction, we would find a preponderance of decisions, to offer or withhold an interpretation, occur at what dynamic systems theory would consider to be phase transitions, and which we see as regressions, splits or projective identifications, however small. By the same token, Beebe and Lachmann have documented in exquisite detail the entrainment of posture, tone, and other interactive aspects of their treatment experiences. This is the essence of the interactive approach, but it is also the case that movement is not back and forth along a linear path, but, as it were, throughout a multi-dimensional space of continually changing phase transitions through state space.

From this point of view, it seems to me that we cannot posit that the self exists as some stable state of affairs that can, even in theory, be attained. Rather, the self needs to be conceptualized as the sum of the available attractor states within phase space through which a process of self-organizing emergence can take place at any given point in time. Thus the self is neither the terminus a quo (away from which development proceeds), as in Fordham, nor the terminus ad quem (toward which development proceeds), as in Kohut. Rather, to think in mathematical terms once again, the self is analogous to the differential equation that defines the state space and the phase transitions of an individual's life pattern.

Clearly, one could argue that by this formulation I have so attenuated our sense of the self as to render it useless as a category. In reply, I can only introduce Jung's definition of the self and try to return to the discussion of myth and situated action with which I began this paper. Jung's definition of the self, which occurs throughout his mature work, is that talk of the self is a way of conceptualizing wholeness (Jung 1959), what in the world of dynamic systems theory we could call a super-ordinate organizing principle that overarches the system of the psyche, and even the system psyche-world. The evident depersonalization of this definition has been one of the major stumbling blocks in the efforts of Jungians and post-Freudians to comprehend one another. In elaborating his notion of the self, Jung argued that the presence of the self is most commonly associated with what he called god-images. But what exactly is a god-image? All too often, god-images are taken as just that, an iconic representation of a deity. But Jung's notion of the image was profoundly contextual, taking into account the entire setting of an organism's life and behaviour (Hogenson 2001). I would therefore argue that for Jung a god image defines the entire dynamic context, implicit and explicit, within which the individual is acting. The god image is, in other words, the phase space of the individual's life organization.

In a slightly different, yet I believe complementary manner, Terrance Deacon remarks, regarding the self:²

Consciousness of self . . . implicitly includes consciousness of other selves, and other consciousnesses can only be represented through the virtual reference created by symbols. The self that is the source of one's experience of intentionality, the self that is judged by itself as well as by others for its moral choices, the self that worries about its impending departure from the world, this self is a symbolic self. It is a final irony that it is the virtual, not actual, reference that symbols provide, which gives rise to this experience of self. This most undeniably real experience is a virtual reality.

(Deacon 1997, p. 452)

Conclusion

What are symbols symbols of? We can now venture a preliminary answer to this question; symbols are not simply representations of states of affairs in the world. In one sense, according to Lacan, at the level of the symbolic, there is no signified. There is only process in the world of the symbolic. Myths and symbols are devices that we use to bootstrap ourselves into the genuinely human world. But by virtue of the symbolic, the human world becomes an infinite world, where the sense of oneself is a constantly emerging sense of potential states and phase transitions that have yet to be traversed. In terms of the relationship between science and the symbolic world, I believe we are on the threshold of an exciting opportunity to in fact explore new paradigms for both psychoanalysis and analytical psychology. However, from the point of view of dynamic systems theory we will have to accommodate ourselves to the notion that many of our theoretical and clinical verities are not as stable as we would like them to be. It may be that we will have to grasp the irony that the psychodynamic depth psychologies are themselves the most dynamic systems we confront. Jung was the first to argue that in an analysis genuinely aimed at transformation the clinician would have to be prepared to undergo his or her own transformation. This was a genuinely dynamic point of view that has recently come to the fore in relational psychoanalysis. Taking it seriously, with all its implications, however, remains the challenge before us.

Translations of Abstract

Cet article pose la question de comment comprendre les symboles en psychologie analytique et en psychanalyse. L’accent est mis sur la tendance récente à utiliser des modèles cognitifs et développementaux dans les deux disciplines, et une revue critique rapide des arguments cognitifs et développementaux est donnée. L’article développe ensuite une argumentation s’appuyant sur la théorie des systèmes dynamiques dans laquelle il n’y a aucun présupposé d’un schème ou d’une structure pré-existante aussi bien dans l’activité de l’esprit que dans l’élaboration des comportements. Dans le modèle des systèmes dynamiques, le soi est vu comme un phénomène d’émergence provenant des schémas dynamiques qui existent dans un système complexe incluant les caractéristiques physiologiques du nouveau-né, les attributions introduisant une intentionnalité du donneur de soin et les ressources culturelles ou symboliques qui constituent l’environnement. Le symbole peut alors être vu comme étant un élément autonome de différentes et importantes manières dans le système dynamique. Sont tirées des conclusions de ce travail ouvrant sur des recherches futures quant à la nature du symbole avec des implications à la fois pour la pratique et pour la théorie en psychologie analytique et en psychanalyse

Diese Arbeit widmet sich der Frage, wie Symbole in Analytischer Psychologie und Psychoanalyse verstanden werden sollten. Die untersuchte Sicht konzentriert sich auf die kürzliche Wendung zu stärker kognitiven und entwicklungspsychologisch bestimmten Modellen in beiden Disziplinen und schildert und kritisiert in Kürze die Argumente aus dem Bereich der Evolution und Kognition. Danach stellt die Arbeit ein Argument vor, das auf der dynamischen Systemtheorie beruht, in der kein Vorbestehen einer Schablone oder Struktur für weder Geist noch Verhalten angenommen wird. Innerhalb des dynamischen Systemmodells wird das Selbst als ein sich entwickelndes Phänomen gesehen, das sich ableitet von den dynamischen Mustern in einem komplexen System, zu dem die physiologischen Charakteristika des Säuglings, die intentionalen Zuschreibungen der versorgenden Person und die kulturellen oder symbolischen Resourcen gehören, welche die Umgebung ausmachen. Das Symbol kann dann als diskretes und in wichtiger Hinsicht autonomes Element im dynamischen System gesehen werden. Es werden Schlußfolgerungen für weitere Forschung über die Natur des Symbols gezogen, mit Implikationen für Theorie und Praxis in der Analytischen Psychologie und Psychoanalyse.

Questo lavoro parla del problema di come dovrebbero essere compresi i simboli nella psicoanalisi e nella psicologia analitica. Il punto di vista in esame focalizza la sua attenzione sul recente volgersi a modelli evolutivi e cognitivi in entrambe le discipline e brevemente rivede e critica le argomentazioni evolutive e cognitiva. Il lavoro presenta poi una discussione basata sulla teoria dei sistemi dinamici in cui non si assume nessun appoggio né struttura preesistente sia per la mente che per il comportamento. All’interno del modello dei sistemi dinamici il Sé viene visto come un fenomeno emergente che deriva dai comportamenti dinamici esistenti in un sistema complesso che include le caratteristiche fisiologiche dell’infante, le attribuzioni intenzionali di chi si prende cura di lui e le risorse culturali o simboliche che costituiscono l’ambiente. Il simbolo può essere visto come un elemento distinto del sistema dinamico, e in modo rilevante autonomo. Le conclusioni sono a favore di un’ulteriore ricerca sulla natura del simbolo con implicazioni che riguardano sia la teoria che la prassi e nella psicologia analitica e nella psicoanalisi.

Este trabajo se pregunta de cómo los símbolos deben ser entendidos en la Psicología analítica y en el psicoanálisis. El punto de vista examinado se enfoca en el reciente viraje hacia los modelos cognitivos y desarrollistas en ambas disciplinas y brevemente revisa y critica los argumentos evolutivos y cognitivos. En el trabajo se asume que un argumento basado en la teoría de los sistemas dinámicos establece que no hay un patrón o estructura preexistente para la mente o la conducta. En el modelo dinámico de sistemas se observa al Self como un fenómeno emergente derivado de los patrones que existen en un sistema complejo el cual incluye las características fisiológicas del infante, las atribuciones intencionales de quien lo cuida y los recursos culturales y simbólicos que constituyen el entorno. El símbolo puede ser visto como un elemento discreto, y en forma importante autónomo, en el sistema dinámico. Las conclusiones determinan la necesidad de nuevas investigaciones sobre la naturaleza del símbolo que tengan implicaciones para ambas, la teoría y la práctica en psicología analítica y psicoanálisis.

Footnotes

• 1

  As part of the presentation of this paper, I showed a very short video that illustrates the nature of situated robotics as a research tool. Readers can view this video clip at http://www.cs.ualberta.cal~kube/research.html where they should click on the short, 5MB, research film. Obviously, if they want to watch the 14MB version they are welcome to, but I used the short version. Readers can also read a discussion of the research that prompted this particular bit of video. I apologize to my Jungian colleagues who have already seen this clip, but it remains the best and most accessible demonstration of simple situated action, as I discuss it in this paper, that I have thus far run across.Regarding phase spaces, I recommend playing around with the demonstration at http://www.serendip.brynmawr.edu/chaos/phase.html. It is a very simple phase space generator, and if you permutate the variable for a while you will see some simple phase transitions. The important thing here is that a system can remain very stable for a long period, and then make an abrupt transition to an entirely new pattern.

• 2

  As I was completing this paper I discovered that Thomas Metzinger, of the Johannes Gutenberg-Universität, Mainz, has advanced a similar position on the virtual nature of self representations (Metzinger 2003). I have not as yet been able to fully assess his position, however.